Study setting and study population
A quasi-experimental study will be conducted in 30 selected privately-owned primary healthcare clinics, and beauty and food establishments in two neighborhoods in Central Singapore. A prior sequential mixed methods study has identified the clinics, beauty and food establishments as the most frequently visited establishments by residents in the community.
Clients (patients/patrons) who visit the participating establishments (10 clinics, 10 beauty establishments, and 10 food establishments) will be invited to participate in the study by scanning the QR code on recruitment posters placed within the establishment premises. Upon scanning the QR code, participants will proceed to complete a pre-intervention questionnaire survey (#1) on an online platform after going through the study information sheet. Upon completion of the survey #1, participants will be guided through embedded instructions to click on a link to access a health education video. After viewing the video, the participants will be led to a post-intervention survey (#2) via an embedded link. Before submitting survey #2, contact details (mobile number) and the preferred payment method will be collected for reimbursement purposes and future follow-up surveys. The enrolled participants will be re-contacted by the study team via a mobile messaging service (such as WhatsApp), and will be guided through embedded instructions to complete three more surveys (#3, #4 & #5) at 1-, 3-, and 6-month post-intervention to assess knowledge on antibiotic use and AMR and antibiotic behaviors.
Participants must be aged 21 years and above, own a smartphone, and not been enrolled in the study at another participating site.
Intervention
A concise two-minute animated video providing educational information on the appropriate use of antibiotics and AMR will be made available through survey #1, as mentioned above, at the participating clinics, beauty and food establishments. The video’s content and format were informed by findings from a previous community survey conducted in the same two neighborhoods.
Outcome measures
The primary outcomes of interest are: (1) Proportion of participants with good knowledge of antibiotic use based on three questions from the WHO’s Antibiotic Resistance Multi-country Public Awareness Survey questionnaire [12, 18]; (2) Proportion of participants with good knowledge of AMR based on eight questions from the WHO’s Antibiotic Resistance Multi-country Public Awareness Survey questionnaire [12, 18]; (3) Proportion of participants who use antibiotics appropriately based on six questions adapted from the US Centers for Disease Control and Prevention’s advisory on appropriate antibiotic use [12, 19]. Participants’ knowledge and practices will be measured longitudinally at five time points: (1) Pre-intervention; (2) Immediate post-intervention; (3) One-month post-intervention; (4) Three-month post-intervention; and (5) Six-month post-intervention.
The secondary outcomes of interest include the reach of the intervention (the number and socio-demographic profiles of participants recruited by the different establishment types) and implementation fidelity (the extent to which the participating sites properly display the study poster and degree to which staff members of the participating sites actively direct their clients’ attention to the poster), using the RE-AIM framework [20].
Sample size
At the end of the one-year recruitment period, a total of 3300 patients/patrons of the participating healthcare clinics, beauty and food establishments will have participated in the study. Based on findings from a prior nationally-representative population-based survey, 59% of Singaporeans correctly answered all three knowledge statements on antibiotic use and 3% correctly answered all eight knowledge statements on AMR [12]. Assuming that the educational intervention facilitated by primary healthcare clinics can result in at least 10% of participants having a good knowledge of AMR at 6-month post-intervention, and beauty and food establishments in at least 6% of participants having a good knowledge of AMR at the same time-point, a sample size of 771 per establishment type would be adequate to detect the difference with 80% power at an alpha level of 0.05. Factoring in a loss-to-follow up rate of 30%, 1100 participants from each establishment type (clinic, beauty, and food) will need to be recruited. A sample size of 1100 participants per establishment type would be adequate to detect most meaningful differences with odds ratios of > 1.25 when the proportion of participants in primary healthcare clinics with good knowledge of antibiotic use or AMR at 6-month post-intervention is below 30%, and differences with odds ratios < 1.25 if the proportion of participants in primary healthcare clinics with good knowledge of antibiotic use or AMR at 6-month post-intervention is > 30%, with a power of 80% and at an alpha level of 0.05.
Data analysis
The unit of analysis is the patient/patron of the participating sites. Descriptive analysis such as means (standard deviations) and medians (lower and upper quartiles) for continuous variables and proportions for categorical variables including the outcomes of interest will be performed, and compared between participants who received the educational intervention from primary care clinics and those who received the intervention from beauty and food establishments respectively. Multiple logistic regression models will be constructed to assess the comparative effectiveness of educational intervention facilitated by primary care clinics, beauty and food establishments. Furthermore, mixed effects logistic regression analysis will be conducted to assess knowledge retention and behavior change by participants over the 6-month period.